Determinants of non-compliance with the pregnancy prevention program in women initiating oral isotretinoin therapy: a cohort study from the French nationwide claims database

Background: Due to the teratogenic effect of isotretinoin, women of childbearing age treated by this acne therapy must comply with the Pregnancy Prevention Program (PPP). However, in France, the number of pregnancies exposed to oral isotretinoin is still not negligible and the determinants of PPP non-compliance are still under-researched.

Objectives: To assess the determinants of the PPP non-compliance among French women initiating oral isotretinoin therapy.

Methods: A new-user cohort study was conducted among women aged 11-50 years between 2014 and 2021 from the National Health Data System (SNDS). Pregnancy tests (PT) were identified from reimbursed laboratory tests of serum or urine βHCG. The PPP non-compliance was assessed at initiation and renewals, and defined by both the absence of PT within 3 days before the prescription and the prescription within 7 days before the dispensation. Determinants were temporal factors (study period and season), age, contraception (intrauterine device or implant), prescriber type, socioeconomic factors (complementary health insurance coverage [CMU-c] and deprivation of the residence area), and access to care for general practitioners, dermatologists, pharmacies and laboratories. Log-Binomial and Poisson regression models were expressed in adjusted relative risks (aRRi at initiation and aRRr at renewals) and 95% confidence intervals (95% CI).

Results: Among 272,723 treatment initiations, 39.0% did not respect the PPP at initiation and 28.4% at renewals. At initiation and renewals, the risk of non-compliance increased i) in summer compared to fall (p < 0.001), ii) among women aged 35-50 years compared to those aged 11-24 years (aRRi 1.33, 95% CI 1.31-1.35; aRRr 1.23, 95% CI 1.21-1.24), iii) among women living in the most area-level deprivation (aRRi 1.07, 95% CI 1.05-1.08; aRRr 1.13, 95% CI 1.12-1.15), iv) when the prescriber was a general practitioner in comparison to a dermatologist (aRRi 1.42, 95% CI 1.40-1.44; aRRr 1.43, 95% CI 1.41-1.45), and v) among women living in areas with the highest access to pharmacies and laboratories (p≤0.004). The risk of non-compliance was also higher among women with a CMU-c (aRRr 1.03, 95% CI 1.02-1.04), only at renewals. Conversely, the risk of non-compliance decreased during the study period (p < 0.001), among women using contraception (p < 0.001) and among women living in areas with high access to general practitioners, dermatologists, and laboratories (p≤0.004).

Conclusions: This large cohort has highlighted that temporal and socioeconomic determinants as well as age and access to care could be taken into account to improve PPP recommendations and tailor them to specific subpopulations.